RFR: 8310190: C2 SuperWord: AlignVector is broken, generates misaligned packs [v57]

[Christian Hagedorn]

On Thu, 21 Dec 2023 15:40:16 GMT, Emanuel Peter <epeter at openjdk.org> wrote:

>> I want to push this in JDK23.
>> After this fix here, I'm doing [this refactoring](https://github.com/openjdk/jdk/pull/16620).
>> 
>> To calm your nerves: most of the changes are in auto-generated tests, and tests in general.
>> 
>> **Context**
>> 
>> `-XX:+AlignVector` ensures that SuperWord only creates LoadVector and StoreVector that can be memory aligned. This is achieved by iterating in the pre-loop until we reach the alignment boundary, then we can start the main loop properly aligned. However, this is not possible in all cases, sometimes some memory accesses cannot be guaranteed to be aligned, and we need to reject vectorization (at least partially, for some of the packs).
>> 
>> Alignment is split into two tasks:
>>  - Alignment Correctness Checks: only relevant if `-XX:+AlignVector`. Need to reject vectorization if alignment is not possible. We must check if the address of the vector load/store is aligned with (divisible by) `ObjectAlignmentInBytes`.
>>  - Alignment by adjusting pre-loop limit: alignment is desirable even if `-XX:-AlignVector`. We would like to align the vectors with their vector width.
>> 
>> **Problem**
>> 
>> I have recently found a bug with our AlignVector [JDK-8310190](https://bugs.openjdk.org/browse/JDK-8310190).
>> In that bug, we perform a misaligned memory vector access, which results in a `SIGBUS` on an ARM32 machine.
>> Thanks @fg1417 for confirming this!
>> Hence, we need to fix the alignment correctness checks.
>> 
>> While working on this task, I also found some bugs in the "alignment by adjusting pre-loop limit": there were cases where it did not align the vectors correctly.
>> 
>> **Problem Details**
>> 
>> Reproducer:
>> 
>> 
>>     static void test(short[] a, short[] b, short mask) {
>>         for (int i = 0; i < RANGE; i+=8) {
>>             // Problematic for AlignVector
>>             b[i+0] = (short)(a[i+0] & mask); // best_memref, align 0
>> 
>>             b[i+3] = (short)(a[i+3] & mask); // pack at offset 6 bytes
>>             b[i+4] = (short)(a[i+4] & mask);
>>             b[i+5] = (short)(a[i+5] & mask);
>>             b[i+6] = (short)(a[i+6] & mask);
>>         }
>>     }
>> 
>> 
>> During `SuperWord::find_adjacent_refs` we used to check if the references are expected to be aligned. For that, we look at each "group" of references (eg all `LoadS`) and take the reference with the lowest offset. For that chosen reference, we check if it is alignable. If yes, we accept all references of that group, if no we reject all.
>> 
>> This is problemati...
>
> Emanuel Peter has updated the pull request incrementally with one additional commit since the last revision:
> 
>   Apply suggestions from code review by Christian
>   
>   Co-authored-by: Christian Hagedorn <christian.hagedorn at oracle.com>

Marked as reviewed by chagedorn (Reviewer).

Thanks a lot Emanuel for all the discussions and for addressing all my comments online and offline :-)

It looks very good now and it's easy to follow the logic. The proofs are great and really helpful to better understand the (rather simple in the end) code for proving and calculating the alignment solutions. Thanks for putting the extra effort in here.

I will have another complete look at the entire PR in the new year. But I think it looks good!

-------------

PR Review: https://git.openjdk.org/jdk/pull/14785#pullrequestreview-1794654777
PR Comment: https://git.openjdk.org/jdk/pull/14785#issuecomment-1867826841